Nano Banana Pro (Gemini 3 Pro Image Preview)
VerifiedGoogle's preview model for advanced image and text tasks.
About Nano Banana Pro (Gemini 3 Pro Image Preview)
Designed as a preview release, this model emphasizes image processing alongside text inputs. Google built it as a non-open-weight system with no public parameter count disclosed. Its architecture prioritizes multimodal handling for preview evaluation.
Strengths include a substantial context window that accommodates extended image-text interactions. The closed nature ensures controlled access during the preview phase. Users benefit from focused image modality support without open-weight distribution.
Typical usage involves testing image analysis and generation workflows in research or development settings. Developers integrate it for multimodal experiments requiring large context. It suits scenarios where Google’s preview tools are needed for image-centric applications.
Capabilities
Best for
Detailed analysis of complex images in extended documents
Processes up to 65536 tokens of multimodal input to deliver grounded text descriptions and reasoning over visuals embedded in long contexts.
Visual question answering on intricate scenes
Combines image understanding with multimodal text-vision reasoning to answer queries about visual content accurately.
Generating explanations from visual data in research workflows
Supports long-context multimodal inputs for producing detailed image descriptions and text generation directly tied to analyzed visuals.
Strengths & limitations
Strengths
- +Strong image-text integration
- +Handles complex visual queries
- +Extended context for scene or document analysis
- +Preview access to advanced Gemini vision features
Limitations
- –Restricted to image and text modalities
- –65k token context limit
- –Preview version may lack full stability or features
Cost calculator
Estimate what Nano Banana Pro (Gemini 3 Pro Image Preview) would cost for your usage.
Based on Nano Banana Pro (Gemini 3 Pro Image Preview)'s $2.00/1M input · $12.00/1M output. Estimate only — actual cost varies by provider and caching.
Quick start
OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. Only the model slug changes between models.
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://openrouter.ai/api/v1",
apiKey: process.env.OPENROUTER_API_KEY,
});
const completion = await client.chat.completions.create({
model: "google/gemini-3-pro-image-preview",
messages: [{ role: "user", content: "Hello!" }],
});
console.log(completion.choices[0].message.content);Model slug: google/gemini-3-pro-image-preview
Editor's verdict
Nano Banana Pro (Gemini 3 Pro Image Preview) is Google's proprietary image models with a 66K-token context window.
At $12.00 per 1M output tokens, it is premium-priced for its class.
It is available through Google's API and aggregators like OpenRouter.
Best suited to strong image-text integration and handles complex visual queries.
Frequently asked questions
The model handles 65536 tokens for long-context multimodal inputs combining text and images.
User reviews
Real, verified reviews from the community shape this model's rating.
Loading reviews…
Similar models
Other image models worth comparing.